Comparative Effectiveness Review Number 100 Effective Health Care Program

Assessment and Management of Chronic Cough Executive Summary

Background Effective Health Care Program In the United States, cough is the most common complaint for which patients The Effective Health Care Program seek medical attention and is the second was initiated in 2005 to provide most common reason for a general valid evidence about the comparative medical examination, accounting for effectiveness of different medical more than 26 million office visits interventions. The object is to help annually.1 Cough often results from consumers, health care providers, an acute, self-limited, viral upper and others in making informed respiratory tract infection; however, choices among treatment alternatives. there are multiple causes of cough Through its Comparative Effectiveness beyond this, including both respiratory Reviews, the program supports tract and nonrespiratory tract-related systematic appraisals of existing etiologies. Cough that lasts more than scientific evidence regarding 4 weeks in children younger than treatments for high-priority health 14 years of age or more than 8 weeks conditions. It also promotes and in adolescents and adults 14 years of generates new scientific evidence by age and older is considered to be identifying gaps in existing scientific chronic by the American College of evidence and supporting new research. Chest Physicians (ACCP).2,3 Cough The program puts special emphasis serves a potentially beneficial purpose on translating findings into a variety by clearing the airways of excessive of useful formats for different mucus, irritants, or abnormal substances stakeholders, including consumers. such as edema fluid or pus. But while The full report and this summary are cough may serve a useful function, it available at www.effectivehealthcare. can also lead to a variety of problems, ahrq.gov/reports/final.cfm. including exhaustion (57%), feeling self-conscious (55%), insomnia (45%), changes in lifestyle (45%), musculoskeletal pain (45%), hoarseness (43%), excessive setting of chronic versus acute cough. As a perspiration (42%), and urinary consequence, chronic cough is responsible incontinence (39%).4 These problems for up to 38 percent of pulmonary are more likely to be prominent in the outpatient visits.5,6

Effective Health Care 1 To effectively assess cough and monitor response to This position is supported by the American Academy of treatment, it is essential to have valid measurement tools. Pediatrics. Currently there are many different tools used to assess The diagnosis and management of cough has been cough frequency and severity, including quality-of-life the subject of several guideline efforts, two aimed at questionnaires, visual analog scales, electronic recordings, assessment of cough in adults,9,10 and one focused on and human counts. It is important to determine whether children.11 Guidelines from ACCP, last updated in 2006, the tools currently in use accurately assess cough and are the most comprehensive resource and will be the response to treatment. While no universally accepted gold subject of a future update.10 standard exists for comparison, data regarding the validity, consistency, reliability, and responsiveness of these tools Identifying the underlying etiology is the most important are needed. The purpose of this review is to evaluate the step in the successful management of chronic cough.10 If, effectiveness of instruments to evaluate cough and the however, no cause can be identified, or if treatment of the comparative effectiveness of treatments for the symptom underlying etiology fails to resolve the cough, then the of cough in patients with either unexplained or refractory cough may be treated symptomatically. In the majority chronic cough. of cases, symptomatic treatment consists of antitussive therapy to decrease cough frequency and severity. In patients with no identifiable cause of cough Antitussive treatments vary in mechanism of action. (unexplained or idiopathic) or no response to specific Nonspecific antitussives such as and treatment (unresponsive, refractory, or intractable), appear to act in the brain stem to reduce the cough chronic cough poses a particularly challenging problem. reflex. Other nonspecific antitussives, such as , The differential diagnosis for chronic cough has a act to anesthetize respiratory passages and thus reduce different list of etiologies compared with acute cough. the stimulus to cough. Other agents aim to decrease the Treatment for chronic cough contrasts with acute cough volume of respiratory tract secretions and thus the need in that acute cough treatment may focus on curing to cough. These latter antitussive agents are also used to the underlying etiology (e.g., bacterial bronchitis or treat certain common underlying etiologies and include pneumonia) or suppressing symptoms for the short period antihistamines, corticosteroids, antibiotics, decongestants, of time needed for the etiology to resolve spontaneously and mast cell stabilizers. Nonpharmacological antitussives (e.g., viral etiologies). Cough becomes chronic if it are few but may include, for example, honey. Recently, persists, often due to an underlying etiology that is speech therapy interventions have been used to treat difficult to diagnose or treat. Therefore, treatments for chronic cough in patients suspected of upper airway cough may have differential effectiveness depending on hypersensitivity.12 whether the cough is acute versus chronic. Side effects of medication may also become more salient in the setting In a limited number of situations where cough provides of chronic cough given that treatment duration is longer, a useful function (such as in bronchiectasis, pneumonia, allowing more opportunity for side effects to occur. or atelectasis), protussive therapy may be used in an Chronic cough also differs from acute cough in that quality attempt to increase cough effectiveness without increasing of life may be affected more severely and in different ways its frequency. Protussive treatments aim to change the than with acute cough. Recent studies from the United characteristics of mucus in such a way that it can be Kingdom, United States, and Japan evaluating patients cleared more effectively by mucociliary action or cough. with chronic cough have estimated that up to 46 percent Such effective clearing can subsequently lessen the of patients have idiopathic cough despite a thorough severity and frequency of a patient’s cough. Protussive diagnostic investigation.7 pharmacological agents include expectorants, mucolytics, and mucus-modifying agents. Examples of these include The management of nonspecific acute or chronic cough , hypertonic saline, and . In in young children can be especially difficult because addition, physical maneuvers such as chest physical of the risks associated with pharmacotherapy. In 2008, therapy, flutter valves, or pneumatic jackets may be used, manufactures voluntarily removed over-the-counter infant especially in patients with respiratory muscle weakness. (<2 years of age) cough and cold products (e.g., those containing ephedrine, pseudoephedrine, phenylephrine, Scope and Key Questions , brompheniramine, or chlorpheniramine) because of many reports of serious adverse events. This comparative effectiveness review (CER) was funded Later that year, manufactures relabeled cough and cold by the Agency for Healthcare Research and Quality products to warn against use in children <4 years of age.8 (AHRQ) and is designed to evaluate the comparative

2 effectiveness of measurement tools for assessing cough b. In patients with refractory cough with a known and of symptomatic treatments for chronic cough. underlying etiology With input from our Key Informants, we constructed Key Figure A depicts the KQs within the context of the Questions (KQs) using the general approach of specifying PICOTS. the population of interest, the interventions, comparators, outcomes, timing of outcomes, and settings (PICOTS). Methods The KQs considered in this review were: The methods for this CER follow those suggested KQ 1: In adults and adolescents (≥14 years of age) and in the AHRQ Methods Guide for Effectiveness and children (<14 years of age), what is the comparative Comparative Effectiveness Reviews (hereafter referred to diagnostic accuracy, therapeutic efficacy, and patient as the Methods Guide)13 and Methods Guide for Medical outcome efficacy of instruments used to assess cough? Test Reviews (hereafter referred to as the Medical Test KQ 2: In adults and adolescents (≥14 years of age) and Guide).14 children (<14 years of age), what are the comparative safety and effectiveness of nonspecific (or symptomatic) Input From Stakeholders therapies to treat patients with chronic cough? During the topic refinement stage, we solicited input from a. In patients with unexplained chronic cough Key Informants representing clinicians (adult and pediatric

Figure A. Analytic framework

Adverse Events • Sleep disturbance • Allergic reaction • Drowsiness Patients With Cough • Headache • Chest pain Adults and adolescents • Dizziness (≥14 years of age) • Rash KQ 2 • Others Children (<14 years of age) Final Outcomes Antitussive • Cough symptoms Underlying Etiologies therapies • Cough severity Patients with • Cough frequency • Asthma unexplained or Protussive KQ 2 • Complications related • GERD refractory therapies to cough • Upper airway cough chronic cough • Functional status syndrome Nonpharmacologic • HRQOL • Tobacco use therapies • Health care utilization • ACE inhibitor user and costs • Pulmonary infection • Bronchitis • Cystic fibrosis Diagnostic Accuracy • Others • Unexplained • Sensitivity, specificity, PPV, NPV, reliability, responsiveness, feasibility, validity Instruments Therapeutic Efficacy • Health-related quality of life instruments • Change in clinical practice • Visual analog scales • Impact on patient and provider • Objective cough counting decisionmaking • Tussigenic challenge KQ 1 Patient Outcome Efficacy • Exhaled nitric oxide • Others • Acceptability • HRQOL • Chest pain, depression, anxiety

ACE = angiotensin-converting enzyme; GERD = gastroesophageal reflux disease; HRQOL = health-related quality of life; KQ = Key Question; NPV = negative predictive value; PPV = positive predictive value

3 pulmonology, otolaryngology, school nursing, respiratory with cough of any duration and considering the following medicine, primary care), patients, scientific experts, and outcomes: diagnostic accuracy (e.g., sensitivity, payers, to help define the KQs. The KQs were then posted validity, reliability, among others); therapeutic efficacy for public comment in September 2011 for 4 weeks, (e.g., impact on patient or provider decisionmaking); and and the comments received were considered in the patient outcome efficacy (e.g., acceptability, quality of development of the research protocol. We next convened life). For KQ 2, the search focused on English-language, the Technical Expert Panel (TEP) comprising clinical, prospective (randomized controlled trial [RCT] or cohort content, and methodological experts to provide input in studies), comparative assessments of pharmacological defining populations, interventions, comparisons, and and nonpharmacological therapies aimed at treating outcomes, and in identifying particular studies or databases the symptom of cough in patients with chronic cough, to search. The Key Informants and members of the TEP in particular, patients with unexplained chronic cough were required to disclose any financial conflicts of interest or refractory cough of known etiology. We accepted as greater than $10,000 and any other relevant business or chronic any cough described as such, or that exceeded professional conflicts. Any potential conflicts of interest 8 weeks in adults and adolescents or 4 weeks in children were balanced or mitigated. Neither Key Informants nor ≤14 years of age. Because determination of whether members of the TEP performed analysis of any kind, nor an individual’s chronic cough was truly unexplained did any of them contribute to the writing of this report. or refractory was often difficult or impossible given available descriptions in the published article, we did not Literature Search Strategy exclude articles based on diagnostic evaluation or empiric To identify the relevant published literature, we searched therapeutic trials, but rather described such information PubMed®, Embase®, and the Cochrane Database of in an attempt to infer to what extent study populations Systematic Reviews (CDSR; last search date for all three could be considered unexplained or refractory according to sources June 4, 2012). Where possible, we used existing current criteria. Articles were excluded if the therapy was validated search filters (such as the Clinical Queries Filters directed at an underlying etiology rather than the symptom in PubMed). An experienced search librarian guided all of cough, if cough resulted from invasive respiratory searches. We supplemented the electronic searches with a tract instrumentation, or if the intervention tested manual search of references from a set of key primary and was not available in the United States. The following systematic review articles. All citations were imported into outcomes were considered: cough symptoms and severity, an electronic database (EndNote® X4; Thomson Reuters, complications related to coughing, functional status, Philadelphia, PA). health-related quality of life, health care utilization and We used several approaches to identify relevant grey costs, and adverse effects of therapy. literature, including a request for scientific information Study Selection packets submitted to drug and device manufacturers and a search of U.S. Food and Drug Administration (FDA) Using the prespecified inclusion and exclusion criteria, device registration studies and new drug applications. We titles and abstracts were reviewed independently by also searched study registries and conference abstracts two investigators for potential relevance to the KQs. for relevant articles from completed studies. Grey Articles included by either reviewer underwent full-text literature databases searched included ClinicalTrials.gov screening. At the full-text review stage, paired researchers (July 18, 2012); the World Health Organization (WHO) independently reviewed the articles and indicated a International Clinical Trials Registry Platform Search decision to “include” or “exclude” the article for data Portal (July 18, 2012); and ProQuest COS Conference abstraction. When the two reviewers arrived at different Papers Index (January 18, 2012). decisions about whether to include or exclude an article, they reconciled the difference through review and Inclusion and Exclusion Criteria discussion, or through a third-party arbitrator if needed. Criteria used to screen articles for inclusion/exclusion at Full-text articles meeting our eligibility criteria were both the title-and-abstract and full-text screening stages included for data abstraction. Relevant review articles, are detailed in Table 2 of the main report. For KQ 1, the meta-analyses, and methods articles were flagged for search focused on English-language evaluative studies manual searching of references and cross-referencing that compared qualitative and/or quantitative instruments against the library of citations identified through electronic used to assess cough in patients (inpatients or outpatients) database searching. All screening decisions were made and

4 tracked in a DistillerSR database (Evidence Partners Inc., adherence to well-accepted standard methodologies and Manotick, ON, Canada). adequate reporting.

Data Extraction For studies of diagnostic tests (KQ 1), we used the QUality Assessment tool for Diagnostic Accuracy Studies The research team created data abstraction forms and (QUADAS)-215 to assess quality in four key domains: evidence table templates for each KQ. Based on clinical patient selection, index test(s), reference standard, and and methodological expertise, a pair of investigators flow and timing. The questions in each domain are rated in was assigned to abstract data from each eligible article. terms of risk of bias and concerns regarding applicability, One investigator abstracted the data, and the second with associated signaling questions to help with these bias reviewed the completed abstraction form alongside the and applicability judgments. original article to check for accuracy and completeness. Disagreements were resolved by consensus or by obtaining Data Synthesis a third reviewer’s opinion if consensus could not be We began our data synthesis by summarizing key features reached. of the included studies for each KQ. We designed the data abstraction forms to collect the For KQ 1 we considered the three dimensions of data required to evaluate the specified eligibility criteria (1) cough frequency, (2) cough severity (which might for inclusion in this review, as well as demographic and include quantity and characteristics of sputum, difficulty other data needed for determining outcomes (intermediate, of expectoration, dyspnea, between cough sensations, final, and adverse events outcomes). We gave particular or pain), and (3) cough-specific quality of life (QOL). attention to describing the details of the treatment, We then sought to measure the validity, reliability, and patient characteristics, and study design that were related responsiveness of various instruments used to assess each to outcomes. In addition, we described comparators of these dimensions. For cough frequency, we evaluated carefully, as treatment standards may have changed validity by concurrence with measures of other constructs during the study period. The safety outcomes were framed (e.g., cough severity, cough-specific QOL, tussigenic to help identify adverse events from drug therapies challenge (or cough reflex sensitivity), and exhaled nitrous and nonpharmacological therapies. Data necessary for oxide), and we assessed reliability using intermethod assessing quality and applicability were also abstracted. reliability (e.g., manual cough counts vs. electronic Before the data abstraction form templates were used, they recording device cough counts) and test-retest reliability. were pilot-tested with a sample of included articles and Although we consider cough severity and cough-specific revised as necessary. QOL to be separate dimensions of cough, most of the standardized questionnaires included in this report Quality Assessment of Individual Studies measured aspects of both of these dimensions. Therefore, We evaluated the quality of individual studies using the for the purpose of this report, we considered instruments approach described in the Methods Guide.13 To assess that measured both severity and QOL together to be quality, we used the strategy to (1) classify the study "severity/QOL" instruments. Within this report, we did design, (2) apply predefined criteria for quality and critical not identify any validated instruments that focused purely appraisal, and (3) arrive at a summary judgment of the on cough severity. For these severity/QOL instruments, study’s quality. Criteria of interest for all studies included we evaluated validity by looking at concurrence with similarity of groups at baseline, extent to which outcomes measures of other constructs including cough frequency, were described, blinding of subjects and providers, quality of life, and tussigenic challenge findings. We blinded assessment of the outcome(s), intention-to- assessed reliability by test-retest reliability, as well as treat analysis, differential loss to followup between the internal consistency. We evaluated responsiveness of both compared groups or overall high loss to followup, and frequency and severity/QOL measures by reporting data conflicts of interest. Criteria specific to RCTs included on changes in these measures over time associated with methods of randomization and allocation concealment. For treatment (or no treatment) of cough symptoms or the observational studies, additional elements such as methods underlying etiology of cough. for selection of participants, measurement of interventions/ For KQ 2, we determined the feasibility of completing exposures, addressing any design-specific issues, and a quantitative synthesis (i.e., meta-analysis). Feasibility controlling confounding were considered. We used the depended on the volume of relevant literature, conceptual summary ratings of good, fair, or poor based on the study’s homogeneity of the studies, and completeness of the

5 reporting of results. We considered meta-analysis for this method uses the PICOTS format as a way to comparisons where at least three studies reported the organize information relevant to applicability. The most same outcome. We considered measures of cough important issue with respect to applicability is whether frequency, regardless of the scale used, to be similar the outcomes are different across studies that recruit enough to combine using effect sizes (standardized mean different populations (e.g., age groups, exclusions for differences); similarly, measures of cough severity that comorbidities) or use different methods to implement the used different measurement scales were considered similar interventions of interest; that is, important characteristics enough to combine using effect sizes. are those that affect baseline (control-group) rates of When a meta-analysis was appropriate, we used random- events, intervention-group rates of events, or both. We effects models to quantitatively synthesize the available used checklists to guide the assessment of applicability. evidence using Comprehensive Meta-Analysis software We used these data to evaluate the applicability to clinical (Version 2; Biostat, Englewood, NJ). We tested for practice, paying special attention to study eligibility heterogeneity using graphical displays and test statistics criteria, demographic features of the enrolled population (Q and I2 statistics). We present summary estimates, in comparison with the target population, characteristics standard errors, and confidence intervals in our data of the intervention used in comparison with care models synthesis. currently in use, and clinical relevance and timing of the outcome measures. We summarized issues of applicability We supplemented the meta-analysis of direct comparisons qualitatively. with a mixed treatment meta-analysis that incorporated data from placebo comparisons and head-to-head Results comparisons, including multi-armed trials (i.e., trials that included more than one comparison). The general strategy Figure B depicts the flow of articles through the for analysis was to construct a random-effects model that literature search and screening process. Searches of was comparable to the standard random-effects models PubMed®, Embase®, and CDSR yielded 21,860 citations, used in the meta-analysis of effect sizes. This model, 6,504 of which were duplicate citations. Manual searching which was fitted using SAS® PROC NLMIXED (2009; identified 75 additional citations, for a total of SAS Institute Inc., Cary, NC), estimated the effect sizes 15,431 citations. After applying inclusion/exclusion (relative to placebo) for each treatment. criteria at the title-and-abstract level, 833 full-text articles were retrieved and screened. Of these, 718 were excluded Strength of the Body of Evidence at the full-text screening stage, leaving 115 articles for data We rated the strength of evidence for each KQ and abstraction. Overall, we included 121 studies represented outcome using the general approach described in the by these 115 publications: 78 studies were relevant to Methods Guide.13,16 and Medical Test Guide.14 In brief, KQ 1, 48 to KQ 2 (5 studies were relevant to both KQs). the approach requires assessment of four domains: risk Studies were conducted in Europe (54%); the United of bias, consistency, directness, and precision. Additional States or Canada (23%); Australia or New Zealand (11%); domains were used when appropriate: coherence, Asia (8%); and other locations (8%). Nineteen studies in dose-response association, impact of plausible residual KQ 1 (23%) and 3 studies in KQ 2 (6%) included children. confounders, strength of association (magnitude of Forty-five studies (37%) were published before 2000. effect), and publication bias. These domains were No additional information was found through our grey considered qualitatively, and a summary rating of “high,” literature search. “moderate,” or “low” strength of evidence was assigned Key Question 1. Instruments Used after discussion by two reviewers. In some cases, high, moderate, or low ratings were impossible or imprudent To Assess Cough to make, for example, when no evidence was available or Key points from the Results chapter are: when evidence on the outcome was too weak, sparse, or inconsistent to permit any conclusion to be drawn. In these • Electronic recording devices are accurate for assessing situations, a grade of “insufficient” was assigned. cough frequency, but they show variable correlation with instruments that measure other dimensions of Applicability cough. We assessed applicability across our KQs using the • The Leicester Cough Questionnaire (LCQ) and the method described in the Methods Guide.13,17 In brief, Cough-specific Quality of Life Questionnaire (CQLQ)

6 Figure B. Literature flow diagram

21,860 citations identified by literature search: PubMed: 10,879 6,504 duplicates Embase: 10,656 Cochrane: 325

Manual searching: 75

15,431 citations identified

14,598 abstracts excluded

833 passed abstract screening 718 articles excluded: – Full-text not available: 24 – Non-English: 8 – Study population is not human: 3 – Not an evaluation study (KQ 1), RCT (KQ 1 or KQ 2), cohort study (KQ 1 or KQ 2): 161 – Study population does not have enough cough (KQ 1) or chronic 115 articles cough (KQ 2): 103 representing 121 unique studies – Study population does not have enough chronic cough of passed full-text screening unknown etiology or refractory chronic cough of known etiology, or has cough resulting from invasive respiratory tract instrumentation (e.g., ventilator dependent, tracheostomy, endotracheal intubation; KQ 2 only): 73 – No intervention of interest or the intervention is intended to

a treat the underlying etiology: 100 Data abstracted for 121 studies: – Did not include outcomes of interest: 215 KQ 1: 78 studies – No comparator: 31 KQ 2: 48 studies

KQ = Key Question; RCT = randomized controlled trial aFive studies were relevant to both KQ 1 and KQ 2.

are the most widely studied cough-specific quality- • Emerging data support the responsiveness of recording of-life questionnaires in adult populations. Both have devices, cough-related questionnaires, and tussigenic demonstrated validity and reliability, with emerging challenge tests, but further research is needed to evidence available on responsiveness. accurately estimate the minimally important difference • There is moderate strength of evidence to support (MID) of these assessment instruments. the validity and responsiveness of the Parent Cough- • Although diaries and visual analog scales are based on specific Quality of Life Questionnaire (PC-QOL) in face validity, assess a wide variety of different cough assessing the severity/QOL of cough among children. outcomes, and are widely used both in research and

7 practice, there is little data to validate their accuracy in (42%) enrolled patients with chronic cough of mixed, assessing cough, and what data exist show inconsistent unknown, or unspecified etiology; 18 (23%) enrolled correlations with other cough measurement tools. These patients with acute cough or cough of unspecified duration, tools are usually simple and easy to use, but more data and 27 (35%) focused on specific clinical conditions such are needed to determine their reliability and validity in as chronic bronchitis, asthma, or lung cancer. Fifty-nine assessing cough frequency or severity/QOL. studies included adults and adolescents (≥14 years of • While all of the included studies evaluated aspects of age), 15 included only children (<14 years of age), and the comparative diagnostic accuracy of the various 4 included adults, adolescents, and children. cough measurement tools, none evaluated the Table A summarizes the findings of our review and comparative therapeutic efficacy or patient outcome the strength of evidence16 for the available outcomes efficacy of these tools. of validity, internal consistency, reliability, and Cough can be assessed along several dimensions, the responsiveness for the main instruments. Details about most of important of which may be frequency, severity, the specific components of these ratings (risk of bias, and cough-specific QOL. Cough frequency is objective consistency, directness, and precision) are available in the and relatively easy to measure but may not necessarily main report. We did not identify any studies evaluating correlate with severity or cough-specific QOL, whereas the comparative therapeutic efficacy or patient outcome cough severity and cough-specific QOL may be closely efficacy of these tools; therefore, the current evidence base interrelated. Most of the standardized questionnaires is insufficient for us to draw any conclusions about these included in this report measured aspects of both of these outcomes. latter dimensions. Therefore, for the purpose of this report, Key Question 2. Nonspecific Therapies we considered instruments that measured both severity for Chronic Cough and QOL together to be “severity/QOL” instruments. In this CER we evaluate the available data that support the Key points from the Results chapter are: validity and reliability of instruments to measure one of • A wide variety of pharmaceutical agents have been two dimensions of cough: (1) cough frequency; or (2) the used to treat the symptom of chronic cough, including severity/QOL impact of cough (including assessments of , anesthetic, and nonopioid/nonanesthetic the impact of cough on sleep, work, general well-being, antitussives; expectorant and mucolytic protussives; health-related quality of life, etc.). We also evaluate the antihistamines; antibiotics; inhaled corticosteroids; and available data that support these instruments’ ability to inhaled anticholinergics. measure potentially meaningful clinical change over time (responsiveness). • Patients with unexplained or refractory chronic cough are not well defined as a population in the evidence To be eligible for inclusion in this report, a study had to base, restricting the applicability of many studies. either (1) compare a cough frequency or severity/QOL assessment instrument with one or more cough assessment, • Of the agents reviewed, the opioid and certain health-related quality of life, or clinical change instrument; nonopioid/nonanesthetic antitussives most frequently or (2) report data on changes in the instrument score over demonstrated efficacy for managing the symptom of time in response to treatment for cough or the underlying chronic cough in adults. etiology of the cough. For the purposes of this report, we • There were several important quality limitations in consider tussigenic challenge tests and exhaled nitric oxide the literature, including (1) too few good-quality tests as severity/QOL assessments. studies focusing on chronic cough; (2) relatively short A total of 78 studies met the inclusion criteria for this KQ. durations of followup (3) a diversity of outcomes Of these, 67 (86%) were judged to have a low risk of bias measured across studies, which limited between-study and 11 (14%) were judged to have a high risk of bias. In comparisons; and (4) when similar outcomes were most cases, the funding source was not reported or was assessed across studies, the instruments used were unclear. Seven studies were RCTs, and the remaining diverse and inconsistent, making comparison and 71 were observational studies. A total of 5,927 participants interpretation difficult. were included across studies; sample sizes of individual • Data on nonpharmacological therapies for chronic studies ranged from 1 to 671 subjects. Thirty-three studies cough were sparse.

8 Table A. Summary of strength of evidence (SOE) and effect estimate for KQ 1a

Instrument Validity (Correlation Reliability (Dimension[s] With Other Internal Consistency Assessed) Measures of Cough) (Cronbach's Alpha) Repeatability Responsiveness Leicester Cough Moderate SOE High SOE High SOE Moderate SOE Questionnaire (LCQ) 15 studies; 1,058 subjects 4 studies; 430 subjects 2 studies; 256 subjects 8 studies; 659 subjects (Severity/QOL) Range of r = 0.26–0.93 Range of r = 0.77–0.93 Range of r = 0.86–0.92 Range of ES = 0.84–19.5 Cough-specific Moderate SOE Insufficient SOE Insufficient SOE Moderate SOE Quality of Life 5 studies; 336 subjects 1 study; 184 subjects 1 study; 52 subjects 7 studies; 460 subjects Questionnaire Range of r = 0.24–0.56 Range of r = 0.63–0.92 Range of r = 0.75–0.93 Range of MID = 10.6–21.9 (CQLQ) and Adverse Cough Outcome Survey (ACOS)b (Severity/QOL) Parent Cough- Moderate SOE Moderate SOE Insufficient SOE Moderate SOE specific Quality of 4 studies; 593 subjects 3 studies; 247 subjects 1 study; 43 subjects 3 studies; 247 subjects Life questionnaire Range of r = 0.01–0.70 Range of r = 0.56–0.91 Ranger of r = 0.40–0.51 Range of ES = 0.32–0.41 (PC-QOL) (Severity/QOL) Electronic recording High SOE NA Moderate SOE Insufficient SOE devices 17 studies; 546 subjects 5 studies; 185 subjects 1 study; 67 subjects (Frequency) Range of r = 0.89–0.99 Range of r = 0.8–1.0 Detected change with treatment Visual analog scales Insufficient SOE NA NA Insufficient SOE (Severity/QOL) 9 studies; 410 subjects 1 study; 21 subjects No summary measure Sensitivity of 0.81–0.95 for detecting clinically important change ACOS = Adverse Cough Outcome Survey; CQLQ = Cough-specific Quality of Life Questionnaire; ES = effect size; KQ = Key Question; LCQ = Leicester Cough Questionnaire; MID = minimal important difference; NA = not applicable; PC-QOL = Parent Cough-specific Quality-of-Life questionnaire; r = correlation coefficient; SOE = strength of evidence aAll strength of evidence ratings of “Insufficient” or “NA” (not applicable) are shaded. bThe ACOS has been revised and replaced by the CQLQ.

• Studies evaluating management of unidentified or A total of 2,923 participants were included across trials; refractory chronic cough in children are extremely sample sizes were relatively small, ranging from 8 to limited. 214 participants. Duration of followup was relatively short • All preparations appeared to be well-tolerated, but side in most studies, ranging from 1 hour to 115 days. Thirty- effects and adverse events were uncommonly reported; three (33) studies (69 percent) had a followup duration of underreporting side effects and adverse events could 2 weeks or less. The majority of studies were rated fair in limit the assessment of effectiveness of these drugs. quality (n=29, or 60%); 11 studies were good in quality, and 8 were poor in quality. Fair- and poor-quality studies Sixty-seven comparisons from 48 studies evaluated had the following limitations: limited description of study therapies in patients with chronic cough and met our entry criteria, randomization, and patient population; inclusion criteria. The 48 studies were described in incomplete followup; less valid statistical analyses (not 42 publications. Thirty-three of the 48 studies were intention-to-treat, post hoc subgroup analyses); and/or parallel-group RCTs, and 12 were randomized crossover inadequate reporting of methods and findings. studies. The range of years of publication was 1953 to 2012; 76 percent of the articles were published before A variety of agents were studied and could be 2000. Only three studies were performed in children. broadly categorized into antitussives, protussives,

9 and nonantitussive/nonprotussive agents. (11 comparisons, 16%) and comparisons between Antitussives were further categorized as opiates, antitussive opiates and other antitussives (10 comparisons, anesthetics, nonpharmacological, or “other” types. 15%). Fourteen different class comparisons were evaluated Protussives included expectorants, mucolytics, and by only one or two studies. Only two studies evaluated nonpharmacological therapies. Nonantitussive/ nonpharmacological interventions. nonprotussive pharmacotherapies included antihistamines, antibiotics, anticholinergics, and bronchodilators. Figure C The heterogeneity of the included studies in terms of the represents the various categories of agents and the interventions and comparators, combined with the lack of comparisons among these agents represented in the three or more studies reporting the same outcome where included studies. The 48 studies represented 67 different there were multiple comparisons, precluded us from comparisons within or between treatment classes and performing meta-analyses on almost all outcomes. Even included studies of 59 individual agents. There were when similar outcomes were assessed across studies, the 39 comparisons (58%) with placebo. The most common instruments used were diverse and inconsistent, making class comparisons were between other antitussives comparison and interpretation difficult. Therefore the and placebo (12 comparisons, 18%), followed by evidence from head to head trials is insufficient to draw comparisons between antitussive opiates and placebo conclusions about relative benefit.

Figure C. Overview of intervention class comparisons

2 7

Protussive Antitussive (Expectorants) (Other) Antibiotic

2 12 10 Anticholinergic 2 2 1 Antitussive 11 Protussive 4 (Opiates) (Mucolytic) Placebo 1 2 1 1 1 3 Antihistamine Antitussive 2 1 (Anesthetics) Protussive (Nonpharmacological) Corticosteroid Antitussive 1 Bronchodilator (Nonpharmacological) 1

10 We were, however, able to evaluate the relative effects on therapies, but the evidence is insufficient to determine cough severity for four classes of treatments for chronic relative benefit among these therapies. cough: antitussive opiates, antitussive dextromethorphan, We performed a similar meta-analysis for cough frequency, antitussive moguisteine, and protussive mucolytics. This including 7 studies and 396 patients. Relative to placebo, analysis included 11 studies and 700 patients. Most of the effect of dextromethorphan on cough frequency the studies compared the treatment with placebo, but one was 0.40 (95% CI, 0.18 to 0.85; p=0.0248), the effect compared opiates with dextromethorphan and placebo. of codeine was 0.57 (95% CI, 0.36 to 0.91; p=0.0260), Because each study used a different measure of severity, and the effect of moguisteine was 0.60 (95% CI, 0.31 to we converted all results to effect sizes (standardized 1.17, p=0.1117). Again, the studies showed significant mean differences). Relative to placebo, the effect of heterogeneity (p=0.0231). The effects of dextromethorphan dextromethorphan on cough severity was 0.54 (95% and codeine compared with placebo on cough frequency confidence interval [CI], 0.27 to 0.80; p=0.0008), the effect support a benefit of these therapies, although the estimates of opiates was 0.63 (95% CI, 0.40 to 0.86; p<0.0001), are too imprecise to determine if one is superior to another. the effect of moguisteine was 0.62 (95% CI, 0.04 to The effect of moguisteine was too imprecise to draw 1.16, p=0.0366), and the effect of mucolytics was conclusions about is efficacy. 0.14 (95% CI -0.20 to 0.49; p=0.384). The studies showed significant heterogeneity (p=0.0023). The effects of Tables B and C summarize the strength of evidence for the dextromethorphan, moguisteine, and opiates compared most commonly used classes of therapies and evaluated with placebo on cough severity support a benefit of these outcomes. Details about the specific components of these

Table B. Summary of strength of evidence (SOE) and effect estimate for KQ 2—active treatment comparisonsa Treatment Comparison Cough Severity Cough Frequency Adverse Effects Antitussive (anesthetic) vs. Insufficient SOE Insufficient SOE Insufficient SOE antitussive (opiate) 1 study; 45 subjects 2 studies; 105 subjects 1 study; 60 subjects Imprecise results Imprecise results Imprecise results Antitussive (opiate) vs. Insufficient SOE Insufficient SOE Insufficient SOE antitussive (other) 16 studies; 958 subjects 8 studies; 655 subjects 5 studies/273 subjects Opiates, dextromethor- Both codeine and No summary measure phan, and moguisteine had dextromethorphan had significant effect sizes vs. significant ES vs. placebo placebo in MTM (ranging in MTM, but wide and from 0.54–0.63), but wide overlapping CIs are too and overlapping CIs are imprecise to draw conclusions too imprecise to (determine about relative effectiveness equivalence or noninferiority or) draw conclusions about relative effectiveness Protussive (mucolytic) vs. Insufficient SOE Insufficient SOE Insufficient SOE antitussive (other) 4 studies; 274 subjects 1 study; 24 subjects 0 studies/subjects Mucolytics had much smaller No summary measure effect size vs. placebo, p=NS, in MTM compared with dextromethorphan Protussive (mucolytic) vs. Insufficient SOE Insufficient SOE Insufficient SOE antitussive (opiate) 4 studies; 274 subjects 1 study; 24 subjects 0 studies/subjects Mucolytics had much smaller No summary measure effect size vs. placebo, p=NS, in MTM compared with codeine CI = confidence interval; ES = effect size; KQ = Key Question; MTM = mixed treatment meta-analysis; NS = not statistically significant; SOE = strength of evidence aAll strength of evidence ratings of “Insufficient” are shaded.

11 Table C. Summary of strength of evidence (SOE) and effect estimate for KQ 2—comparisons with placeboa Treatment Comparison Cough Severity Cough Frequency Adverse Effects Codeine/opiates—Antitussive Low SOE Low SOE Insufficient SOE (opiates) vs. placebo 11 studies; 396 subjectsb 7 studies; 700 subjectsb Imprecise results 0.63 (95% CI, 0.40 to 0.86; 0.57 (95% CI, 0.36 to 0.91; p<0.0001), from MTM p=0.0260), from MTM Dextromethorphan— Low SOE Low SOE Insufficient SOE Antitussive (other) vs. placebo 11 studies; 396 subjectsb 7 studies; 700 subjectsb No summary measure 0.54 (95% CI, 0.27 to 0.80; 0.40 (95% CI, 0.18 to 0.85; p=0.0008), from MTM p=0.0248), from MTM Protussive (mucolytic) vs. Insufficient SOE Insufficient SOE Insufficient SOE placebo 11 studies; 396 subjectsb No summary measure No summary measure 0.14 (95% CI -0.20 to 0.49; p=0.384) from MTM Moguisteine—Antitussive Low SOE Insufficient SOE Insufficient SOE (other) vs. placebo 11 studies; 396 subjectsb 7 studies; 700 subjectsb No summary measure 0.62 (95% CI, 0.04 to 1.16, 0.60 (95% CI, 0.31 to 1.17, p=0.0366), from MTM p=0.1117), from MTM CI = confidence interval; KQ = Key Question; MTM = mixed treatment meta-analysis; SOE = strength of evidence aAll strength of evidence ratings of “Insufficient” are shaded. bTotal number of studies/patients from mixed treatment meta-analysis. ratings (risk of bias, consistency, directness, and precision) for assessing cough frequency, but they show variable are available in the main report. Across outcomes and correlation with other cough measurement tools. This comparisons, although the included evidence was from may be because cough frequency is unidimensional, RCTs with an overall low risk of bias, the findings whereas the impact that cough may have on an individual’s were inconsistent; the evidence, when available, was functional status, quality of life, or sense of wellbeing may indirect (i.e., based on mixed treatment meta-analysis); depend on many other factors. Multidimensional quality- and the findings, when available, were imprecise. There of-life assessments such as the LCQ, CQLQ, and other was insufficient evidence to support conclusions about cough-specific instruments may therefore be more useful comparative effectiveness of the interventions for any of than simple cough frequency in assessing meaningful our key outcomes. Evidence for other comparisons was too impact of cough. Visual analog scales, although widely sparse to construct such summary tables. used both in research and practice, have little to no data to validate their accuracy in assessing cough, and inconsistent Discussion correlations with other cough measurement tools. We reviewed 48 studies involving 2,923 patients that Key Findings evaluated nonspecific (or symptomatic) therapies to treat We reviewed 78 studies involving 5,927 patients that patients with chronic cough. Our review found that a evaluated instruments used to assess cough. Our findings wide variety of pharmaceutical agents have been used suggest that selected cough-specific quality-of-life to treat the symptom of chronic cough. The opioid and instruments are valid and reliable for assessing cough. certain nonopioid/nonanesthetic antitussives demonstrated The LCQ and the CQLQ along with its predecessor, the most promise for managing the symptom of chronic the Adverse Cough Outcome Survey [ACOS]), are cough. In particular, codeine (with dose response and the most widely studied cough-specific quality-of-life placebo-controlled data) and dextromethorphan have questionnaires in adults, with several studies showing fair reasonably good data for reducing cough frequency and to moderate correlation with other cough measurement severity. However, due to inconsistency and imprecision of tools such as cough frequency logs, tussigenic results, and small numbers of head-to-head comparisons, challenges, electronic recordings, or other quality-of-life the overall strength of evidence is insufficient to draw firm questionnaires. Electronic recording devices are reliable conclusions about the comparative effectiveness of these

12 agents. Finally, the evidence exploring the effectiveness or refractory chronic cough. More often patients were of treatments in patients with truly unexplained cough selected from persons with chronic cough (of variable was minimal. We considered the vast majority of study duration) with a variety of diseases associated with cough. populations to have unresponsive chronic cough. Only While we tried to apply criteria to improve applicability three studies, including one of morphine, were clearly in (e.g., excluding cystic fibrosis and bronchiectasis), patients with unexplained cough and required subjects to the studies we ultimately included contained more have gone through a diagnostic evaluation to exclude most diversity than we intended. In particular, studies with causes of cough. Interestingly, therapy in each of these mixed etiologies of cough (including, e.g., patients with studies was associated with a reduction in cough severity, tuberculosis or lung cancer) and studies from different eras suggesting that chronic unexplained cough can respond to and geographic locations challenge the usefulness of data nonspecific therapies aimed at the symptom and not the on treatment. The majority of studies took place in Europe, underlying etiology. with 9 in the United Kingdom and 17 in other countries in Europe (total of 54%); only 9 (19%) took place within the Unfortunately, we identified only one study of a currently United States or Canada. available (in the United States) treatment (amoxicillin clavulanate) in children with chronic cough, but the study’s For the studies focusing on the adult population, many applicability was limited in terms of its sample size and the of the drug treatment trials we identified included drugs description of the diagnostic evaluation of cough. Given that are not currently available in the United States the lack of studies on treatment of chronic unexplained (12 studies, 30 percent). The applicability of the included cough in children, it is not surprising that there were no studies was also reduced given the age of much of the data on harms in this population. evidence, and therefore of the corresponding interventions and underlying clinical management of the patients. Applicability Publication dates ranged from 1953 to 2012, with 32 (76%) of the articles being published before 2000. It is reasonable to assume that the utility, performance, Given the changes in both available therapies and the reliability, and validity of cough instruments may differ diagnosis and treatment of underlying etiologies, more between children and adults, between acute and chronic recent studies of contemporary therapies are needed. cough conditions, and between underlying etiologies such as asthma, chronic bronchitis, acute rhinitis, lung Limitations of the Comparative cancer, and chronic refractory cough. More consistent reporting of patient characteristics such as age, underlying Effectiveness Review Process etiology, duration of symptoms and/or illness, overall Our findings have limitations related to the literature medical comorbidity, and prior treatment would facilitate and our approach. Important limitations of the literature evaluations of various cough instruments in important include: (1) few studies exploring the clinical population subgroups. For our analysis of instruments for the of interest (unexplained or refractory chronic cough) and assessment of cough (KQ 1), most of the studies were in specific patient subgroups of interest (e.g., children, conducted in Europe (41 studies, 53%); 32 of these women, immunocompromised patients); (2) variable were conducted exclusively in the United Kingdom. definitions of chronic cough; (3) diverse etiologies of Nineteen (24%) studies were conducted in the United cough that might respond differently to different therapies; States or Canada. Location of study was not, however, (4) incomplete reporting of patient characteristics, study obviously related to design, patient, outcome, or analytical design, or outcomes; (5) small sample sizes and short characteristics. duration of followup; (6) lack of gold standard outcomes By restricting inclusion to trials of patients with to assess efficacy and tolerability; and (7) inconsistent unexplained or refractory cough, we improved the reporting of comparative statistical analyses. In addition, applicability of our findings to this population but also most of the studies were comparatively old, and as such decreased the availability of evidence that could be the evidence base suffers from age because of advances in reviewed. Expanding our evidence to include patients clinical trial methodology, improved diagnostic evaluation with acute cough would have substantially increased the of cough, and development of valid and reliable measures evidence base but greatly reduced the applicability of the for cough and cough-specific quality of life. findings to the unexplained or refractory chronic cough Our review methods also had limitations. Our study was population. Few studies directly reported assembling limited to English-language publications. In addition, patients fitting our intended population of idiopathic even within patients with chronic cough, the target

13 population of patients with unexplained chronic cough identify gaps in evidence and describe why these gaps or refractory chronic cough with a known underlying exist.19 This approach considers PICOTS (population, etiology was difficult to identify. Rarely was a thorough intervention, comparator, outcomes, timing, and setting) to negative diagnostic evaluation performed to assemble identify gaps and classifies gaps as due to (1) insufficient a group with unexplained chronic cough; in the case of or imprecise information, (2) biased information; studies of patients with a known underlying etiology, (3) inconsistency or unknown consistency, and (4) not seldom was previously tried therapy described well the right information. Results are as follows: enough to determine whether patients were treatment- KQ 1—Instruments used to assess cough: refractory. In general, we considered use of a symptomatic treatment in a population with a known underlying • Evidence establishing the responsiveness, validity, etiology to imply refractory cough unless patients were reliability, and consistency of available assessment noted to be treatment-naïve; certain etiologies, however, instruments other than the LCQ and CQLQ, and were considered differently. For example, most studies building on available evidence for the LCQ and CQLQ of cough-variant asthma, a common cause of chronic instruments cough in children, which is usually highly responsive to • Additional validation or measurement studies focusing appropriate asthma management, were excluded. on the pediatric population in addition to the limited It is possible that our a priori definition of chronic cough studies that report on the PC-QOL in childhood (i.e., persisting at least 4 weeks if <14 years • Development and validation of child/patient-completed, of age, or 8 weeks if 14 years or older) was too long and cough-specific quality-of-life instruments (as opposed did not reflect care delivery. However, our decision to to parent/proxy instruments such as the PC-QOL) include studies that described their population as suffering • Feasibility of cough assessment instruments in usual with chronic cough regardless of time cut-off may have care (outside of RCTs or validation studies) mitigated this problem. Focusing on nonspecific or symptomatic treatments to the exclusion of treatments – Insufficient evidence curently exists; could be aimed at specific causes of chronic cough proved more explored through observational studies complicated to implement than we had anticipated. Certain • Uncertainty about the effects of patient self-reporting, therapies that we classified as specific (e.g., antihistamines parent reporting, or provider reporting in use of cough and decongestants for upper airway cough syndrome) assessment tools are sometimes referred to as nonspecific.18 Furthermore, – Insufficient evidence curently exists; could be some other specific treatments were occasionally tested explored through observational studies as nonspecific treatments in populations that did not meet diagnostic criteria for conditions for which the • Incomplete evidence regarding the minimally important specific treatment would be appropriate. Finally, we difference of cough frequency or severity/QOL grouped antitussive and protussive drugs into subsets that instruments sometimes included pharmacologically diverse agents or • Impact of measurement tools on therapeutic efficacy or separate drugs with certain similarities. patient outcome efficacy Research Gaps KQ 2—Nonspecific therapies for chronic cough: We found sufficient evidence to suggest that the LCQ and • Comparative effectiveness of pharmacological CQLQ (for adults) and the PC-QOL (for children) may be therapies in the adult population valid instruments for assessing severity/QOL of cough, – Current evidence is both imprecise and inconsistent. and that electronic recording devices, in general, appear to Additional comparative RCTs of contemporary and be valid assessments of cough frequency compared with available agents are needed. human cough counts. Unfortunately, however, the current evidence base is insufficient to provide conclusive findings • Comparative effectiveness of pharmacological related to the comparative effectiveness of available therapies in the pediatric population therapies for patients with unexplained or refractory – Current evidence is insufficient and does not chronic cough. There are, therefore, numerous areas of reflect available therapies. Comparative RCTs of evidence gaps and areas for potential future research. We contemporary and available agents specific to the used the framework recommended by Robinson et al. to pediatric population are needed.

14 • Comparative effectiveness of nonpharmacological anticholinergics. There were relatively few good-quality therapies in both adult and pediatric populations studies focusing on chronic cough using reliable outcome – Current evidence is insufficient. Comparative RCTs measurements over durations of followup pertinent of contemporary and available agents specific in to chronic cough. The opioid and certain nonopioid/ both adult and pediatric populations are needed. nonanesthetic antitussives most frequently demonstrated efficacy for managing the symptom of chronic cough – Additional RCTs or potentially patient-level meta- compared with placebo, but there were insufficient analyses of existing and future RCTs focusing data to draw conclusions between therapies. Data on on subpopulations of interest including women, nonpharmacological therapies for chronic cough are pregnant women, patients with specific underlying extremely limited, as are data on the management of etiologies, immunocompromised patients, and unidentified or refractory chronic cough in children. patients with a history of substance abuse Our systematic review highlights the clear need for • Comparative effectiveness of available therapues in further studies in patient populations with unexplained impacting health utilization and costs or refractory chronic cough as determined by current – Insufficient evidence curently exists; could be diagnostic and empiric treatment recommendations. explored through observational studies Further, it shows the need for more systematic design and reporting of these studies and assessment of patient- • Comparative effectiveness of available therapies in centered outcomes. impacting cough severity, frequency, and quality of life – Current evidence is both imprecise and inconsistent. References Additional comparative RCTs using standardized 1. Centers for Disease Control and Prevention. National Ambulatory instruments are needed. Medical Care Survey: 2008 Summary Tables. www.cdc.gov/nchs/ ahcd.htm. Conclusions 2. Chang AB, Glomb WB. Guidelines for evaluating chronic cough There is no established gold standard for assessing either in pediatrics: ACCP evidence-based clinical practice guidelines. frequency or severity/QOL of cough, thereby making it Chest. 2006;129(1 Suppl):260S-283S. PMID: 16428719. difficult to quantitatively assess test accuracy for cough 3. Pratter MR. Overview of common causes of chronic cough: instruments. Validity of severity/QOL questionnaires ACCP evidence-based clinical practice guidelines. Chest. was generally demonstrated in the published literature 2006;129(1 Suppl):59S-62S. PMID: 16428693. by correlation with other cough assessment instruments, 4. Irwin RS, Curley FJ. The treatment of cough. A comprehensive whereas validity of cough recording devices was review. Chest. 1991;99(6):1477-84. PMID: 2036833. generally demonstrated using human cough counts as 5. Irwin RS, Corrao WM, Pratter MR. Chronic persistent cough in the reference standard. Reliability of questionnaires was the adult: the spectrum and frequency of causes and successful generally demonstrated by test-retest correlation and by outcome of specific therapy. Am Rev Respir Dis. demonstrating internal consistency. Several instruments, 1981;123(4 Pt 1):413-7. PMID: 7224353. including the LCQ, CQLQ, and the PC-QOL, show good 6. Irwin RS, Curley FJ, French CL. Chronic cough. The spectrum internal consistency but variable correlation with other and frequency of causes, key components of the diagnostic cough measurement tools. This suggests that these tools evaluation, and outcome of specific therapy. Am Rev Respir Dis. 1990;141(3):640-7. PMID: 2178528. may be reliable but demonstrate variable validity. The lack of validated reference tests and the diverse number 7. Chung KF, Pavord ID. Prevalence, pathogenesis, and causes of chronic cough. Lancet. 2008;371(9621):1364-74. of instruments used among treatment evaluations also PMID: 18424325. complicates comparisons across studies. We identified no evidence exploring the impact of cough assessment 8. U.S. Food and Drug Administration. Special Features: An Important FDA Reminder for Parents: Do Not Give Infants instruments on therapeutic efficacy or patient outcome Cough and Cold Products Designed for Older Children.www.fda. efficacy. gov/Drugs/ResourcesForYou/SpecialFeatures/ucm263948.htm. A wide variety of pharmaceutical agents have been used 9. Pratter MR, Brightling CE, Boulet LP, et al. An empiric integrative to treat the symptom of chronic cough, including opioid, approach to the management of cough: ACCP evidence-based anesthetic, and nonopioid/nonanesthetic antitussives; clinical practice guidelines. Chest. 2006;129(1 Suppl):222S-231S. expectorant and mucolytic protussives; antihistamines; PMID: 16428715. antibiotics; inhaled corticosteroids; and inhaled

15 10. Irwin RS, Baumann MH, Bolser DC, et al. Diagnosis and Full Report management of cough executive summary: ACCP evidence-based clinical practice guidelines. Chest. 2006;129(1 Suppl):1S-23S. This executive summary is part of the following document: PMID: 16428686. McCrory DC, Coeytaux RR, Yancy WS Jr., Schmit KM, 11. Finnish Medical Society Duodecim. Chronic cough in a child. Kemper AR, Goode A, Hasselblad V, Heidenfelder BL, In: EBM Guidelines. Evidence-Based Medicine [Internet]. Irvine RJ, Musty MD, Gray R, Sanders GD. Assessment Helsinki, Finland: Wiley Interscience. John Wiley & Sons; 2007. and Management of Chronic Cough. Comparative www.guideline.gov/content.aspx?id=12798. Effectiveness Review No. 100. (Prepared by the Duke 12. Gibson PG, Vertigan AE. Speech pathology for chronic cough: Evidence-based Practice Center under Contract a new approach. Pulm Pharmacol Ther. 2009;22(2):159-62. No. 290-2007-10066-I.) AHRQ Publication No. PMID: 19061964. 13-EHC032-EF. Rockville, MD: Agency for Healthcare 13. Agency for Healthcare Research and Quality. Methods Guide for Research and Quality; January 2013. Errata March 2014. Effectiveness and Comparative Effectiveness Reviews. Rockville, www.effectivehealthcare.ahrq.gov/reports/final.cfm. MD: Agency for Healthcare Research and Quality. www.effectivehealthcare.ahrq.gov/index.cfm/search-for-guides- reviews-and-reports/?pageaction=displayproduct&productid=318. 14. Agency for Healthcare Research and Quality. Methods Guide for Errata Medical Test Reviews. Rockville, MD: Agency for Healthcare Table A in the Executive Summary and Tables 6, 11, and Research and Quality. www.effectivehealthcare.ahrq.gov/index. 12 in the full report have been updated to reflect the cfm/search-for-guides-reviews-and-reports/?productid=558& following changes: pageaction=displayproduct. 15. Whiting PF, Rutjes AW, Westwood ME, et al. QUADAS-2: 1. CQLQ--corrected sample size and correlation A revised tool for the quality assessment of diagnostic accuracy coefficients for French 2002 paper for Internal studies. Ann Intern Med. 2011;155(8):529-536. PMID: 22007046. Consistency. 16. Owens DK, Lohr KN, Atkins D, et al. AHRQ series paper 5: 2. CQLQ--corrected sample size and correlation Grading the strength of a body of evidence when comparing coefficients for French 2002 paper for Repeatability. medical interventions—Agency for Healthcare Research and Quality and the Effective Health-Care Program. J Clin Epidemiol. 3. PC-QOL--added data from Newcombe 2010 study 2010;63(5):513-23. PMID: 19595577. for Repeatability. 17. Atkins D, Chang SM, Gartlehner G, et al. Assessing applicability The text and conclusions remain unchanged. when comparing medical interventions: AHRQ and the Effective Health Care Program. J Clin Epidemiol. 2011;64(11):1198-207. PMID: 21463926. 18. Chang AB. Therapy for cough: where does it fall short? Expert Rev Respir Med. 2011;5(4):503-13. PMID: 21859270. 19. Robinson KA, Saldanha IJ, Mckoy NA. Frameworks for Determining Research Gaps During Systematic Reviews. Methods Future Research Needs Report No. 2. (Prepared by the Johns Hopkins University Evidence-based Practice Center under Contract No. HHSA 290-2007-10061-I.) AHRQ Publication No. 11-EHC043-EF. Rockville, MD: Agency for Healthcare Research and Quality. June 2011. www.effectivehealthcare.ahrq.gov/reports/ final.cfm.

AHRQ Pub. No. 13-EHC032-1 January 2013 16 Errata March 2014